Skip to content

fork default: LCCClassesToReplaceNNMethod = 'nearestWeighted'; require LandR (>= 1.2.0.9005) #1

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
achubaty merged 1 commit into from
Aug 5, 2026
Merged

fork default: LCCClassesToReplaceNNMethod = 'nearestWeighted'; require LandR (>= 1.2.0.9005) #1

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
22 changes: 18 additions & 4 deletions Biomass_borealDataPrep.R
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -10,7 +10,7 @@ defineModule(sim, list(
person(c("Alex", "M."), "Chubaty", email = "achubaty@for-cast.ca", role = c("aut"))
),
childModules = character(0),
version = list(Biomass_borealDataPrep = "1.5.12"),
version = list(Biomass_borealDataPrep = "1.5.13"),
timeframe = as.POSIXlt(c(NA, NA)),
timeunit = "year",
citation = list("citation.bib"),
Expand All @@ -23,7 +23,7 @@ defineModule(sim, list(
"archive", "assertthat", "cli", "data.table", "dplyr", "ggplot2", "httr2",
"merTools", "plyr", "qs2", "rasterVis", "sf", "terra", "googledrive",
"reproducible (>= 2.1.0)", "SpaDES.core (>= 2.1.0)", "SpaDES.tools (>= 2.0.0)",
"PredictiveEcology/LandR@development (>= 1.1.5.9090)",
"PredictiveEcology/LandR@development (>= 1.2.0.9005)",
"PredictiveEcology/pemisc@development"
),
parameters = rbind(
Expand Down Expand Up @@ -137,6 +137,16 @@ defineModule(sim, list(
"Since this is about estimating parameters for growth, it doesn't make any sense to have",
"unique estimates for transient classes in most cases. If no classes are to be replaced, pass",
"`'LCCClassesToReplaceNN' = numeric(0)` when supplying parameters.")),
defineParameter("LCCClassesToReplaceNNMethod", "character", "nearestWeighted", NA, NA,
paste("Passed to `LandR::convertUnwantedLCC()` as its `method` argument, controlling how",
"each `P(sim)$LCCClassesToReplaceNN` pixel picks among the available classes in its",
"neighbourhood. Both options weight the classes by their local abundance and differ",
"only in reproducibility. `'nearestWeighted'` (default) keys the draw on the pixel's",
"ground position, so it is deterministic, needs no seed, and a grid-aligned crop of",
"the study area gives the same answer as the full extent. `'nearestRandom'` draws",
"from the RNG instead, so replicates differ -- but note that `Cache()` does not key",
"on RNG state, so a cached call replays a single draw unless the seed is part of the",
"cache key. See `?LandR::convertUnwantedLCC`.")),
defineParameter("minCoverThreshold", "numeric", 5, 0, 100,
"Pixels with total cover that is equal to or below this number will be omitted from the dataset"),
defineParameter("minRelativeBFunction", "call", quote(LandR::makeMinRelativeB(pixelCohortData)),
Expand Down Expand Up @@ -834,14 +844,18 @@ createBiomass_coreInputs <- function(sim) {
newLCCClasses <- convertUnwantedLCC(
classesToReplace = P(sim)$LCCClassesToReplaceNN,
rstLCC = rstLCCAdj,
availableERC_by_Sp = availableCombinations2
availableERC_by_Sp = availableCombinations2,
method = P(sim)$LCCClassesToReplaceNNMethod
) |>
Cache(userTags = c(cacheTags, "newLCCClasses", "stable"))

## adjust rstLCCAdj so that ecoregionMap will contain the last set of updated LCCClassesToReplaceNN
if (nrow(newLCCClasses)) {
if (!is.null(newLCCClasses$newPossLCC)) {
## LandR versions prior to 1.1.5.9045 will not have this
## LandR versions prior to 1.1.5.9045 do not have this, and 1.2.0.9004 dropped it
## again -- where this guard silently stopped firing, leaving rstLCCAdj (and so
## ecoregionMap) showing the un-replaced classes. reqdPkgs now floors LandR at
## 1.2.0.9005, which restored it.
rstLCCAdj[newLCCClasses$pixelIndex] <- newLCCClasses$newPossLCC
}
}
Expand Down
18 changes: 18 additions & 0 deletions NEWS.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,5 +1,23 @@
Known issues: <https://github.com/PredictiveEcology/Biomass_borealDataPrep/issues>

version 1.5.13
=============

## dependency changes
* requires `LandR (>= 1.2.0.9005)`, which reworked `convertUnwantedLCC()`: it gains a
`method` argument, and it again returns the `newPossLCC` column this module needs to
write replacement classes back into `rstLCCAdj`. That column was absent in LandR
1.2.0.9004, where the `is.null()` guard below silently stopped firing and left
`rstLCCAdj` (hence `ecoregionMap`) showing the un-replaced classes.

## new features
* new parameter `LCCClassesToReplaceNNMethod`, passed to `LandR::convertUnwantedLCC()` as
its `method`. Default `"nearestWeighted"` allocates each `LCCClassesToReplaceNN` pixel a
neighbouring class drawn in proportion to that class's local abundance, keyed on the
pixel's ground position so the result is deterministic without a seed and a grid-aligned
crop of the study area agrees with the full extent. `"nearestRandom"` draws from the RNG
instead, for when replicates should differ.

version 1.5.4
=============

Expand Down